IPCC guidance permits two approaches to estimating carbon change, and they are not stylistic alternatives. They demand different data, fail in different ways, and applied to the same forest will frequently produce different numbers. Knowing which one a methodology requires is a prerequisite to designing the monitoring at all.
The Two Approaches
Stock-change (stock-difference) measures the carbon stock at two points in time and subtracts.
Gain-loss estimates the annual increment from growth and subtracts the annual losses from harvest, mortality and disturbance.
| Stock-change | Gain-loss | |
|---|---|---|
| Primary data | Two inventories | Growth rates and loss activity data |
| Detects | Net result, including unknown causes | Only what is explicitly counted |
| Fails when | Interval is short relative to measurement error | A loss pathway is omitted |
| Best for | Longer intervals, measurable pools | Well-characterised managed systems |
Why Gain-Loss Misses Things
Gain-loss is a bookkeeping method. It counts what you tell it to count.
If a project has an unrecorded loss pathway, such as illegal extraction, undetected mortality, or a small fire, gain loss will not see it. The books balance, but the forest does not. Stock change catches it automatically, because the second inventory simply finds less carbon, whatever the cause
This makes stock-change the more honest method in settings where not all losses are documented, which describes most real landscapes.
Why Stock-Change Fails Over Short Intervals
The opposite weakness. Stock-change measures a difference between two large numbers, each with its own uncertainty.
Over a two-year interval in a mature forest, the genuine change may be a small fraction of the measurement error on either inventory. The result is a difference statistically indistinguishable from zero, or worse, a negative value in a forest that is demonstrably growing.
Gain-loss handles this better because growth rates can be estimated with reasonable precision from increment data, and the method is not differencing two large stocks.
The Practical Hybrid
Most well-designed monitoring systems use both, not because a methodology requires it but because disagreement between them is diagnostic.
If gain loss says the forest gained 3 tC/ha/year and stock change says it gained 1, something is being lost that the books do not record. That gap is information, as it points at an unrecorded pathway, a measurement problem, or a growth model that does not match the site.
Projects that run only one method never see that signal. Projects that run both and reconcile the difference find their own errors before an assessor does.
What This Means When Choosing
Long monitoring intervals, permanent plots, a pool that can be measured directly: stock-change works and is more robust to surprises.
Short intervals, well-documented management, reliable local growth data: gain-loss is more precise and will not produce nonsense results.
The methodology usually specifies one. Where it permits a choice, the deciding question is whether you trust your loss records more than you trust your inventory precision — and for most projects operating in real landscapes, the honest answer favours stock-change.
Turn climate goals into a verified portfolio
TREEO connects the full carbon cycle, encompassing eligibility, simulation, real time monitoring, and registry ready reporting, thereby combining expert consulting with dMRV technology so the evidence exists before anyone asks for it.



